Learning-based automated negotiation between shipper and forwarder

This paper studies an automated negotiation system by means of a learning-based approach. Negotiation between shipper and forwarder is used as an example in which the issues of negotiation are unit shipping price, delay penalty, due date, and shipping quantity. A data ratios method is proposed as the input of the neural network technique to explore the learning in automated negotiation with the negotiation decision functions (NDFs) developed by [Faratin, P., Sierra, C., & Jennings, N.R. (1998). Negotiation Decision Functions for Autonomous Agents. Robotics and Autonomous Systems, 24 (3), 159-182]. The concession tactic and weight of every issue offered by the opponent can be learned from this process exactly. After learning, a trade-off mechanism can be applied to achieve better negotiation result on the distance to Pareto optimal solution. Based on the results of this study, we believe that our findings can provide more insight into agent-based negotiation and can be applied to improve negotiation processes.